Don’t Just Listen to the Numbers – Listen to the Users

Interesting eMarketer article this past week about how marketing managers are still only allocating 5-10% of their budget to online media and how one of the biggest obstacles to expanding that budget is having the right analytics or metrics to measure the offline to online influence.  Marketers rated “Having metrics to properly allocate the mix of traditional and digital media” was rated the highest major problem to overcome before further adopting digital media.  This totally surprises me because the amount of digital data that is available and the ability to measure that data in real time is so readily available – whether it’s GoogleAnaltyics, Omniture, or even Nielsen HomeScan which measures offline to online influence.  In fact, when advising companies on whether they should use Omniture, one of my first warnings is to watch out for DATA OVERLOAD.  In referencing to companies about how to implement Omniture my advice is always the following:

  1. Plan out for a three month implementation – they will tell you one month – to do it right think about your data and the relationships (they call it correlations) want to analyze and build that out in a flow chart.
  2. Know your baseline for success before the implementation – Omniture has very different definitions on sessions,visitors, unique visitors and returning visitors then your marketing or accounting department so conversion rates won’t match but you want to get as close as possible to the baseline.
  3. Really understand all their tools because they have some very powerful tools that can really help you in naming conventions so that your terminology for senior management is presented in a way they can understand and appreciate the data.


What marketers should really understand before they allocate any budget to a marketing tactic is that using data itself is only a two-dimensional function.  It can only infer what is happening on a page based on two things: how much time did it take to complete the action and how many pages did they interact with before they left.  What any analytic data lacks is the emotional state of what is happening at the time of the interaction.  What was the person thinking, feeling or experience during their interaction?  Market reach in its classic form – phone survey, focus group, group session demonstrate – where the end user can give their opinion on the experience and product needs to COMPLIMENT any and all path and time based online data.  This analytic data added to the research data should also be varied by data segmentation audience as well.  Meaning that should understand your data audience clusters whether it’s by age, gender, geographic region, etc then apply the research data on top of that.  Now, to be clear – market research data isn’t an ongoing process, it is something that has to be done in conjunction with design changes, product launches, and sometimes with new media tactics – ad campaigns to gather a true read on what you should expect from your analytics data.

Originally posted on Big Couch Media